Hugging Face Trending Papers

When Models Edit Too Much: On the Fidelity of Minimal Code Edits

The paper investigates the problem of over‑editing by large language models (LLMs) when repairing code, showing that even state‑of‑the‑art models like GPT‑5.5 often rewrite more code than necessary. Using a benchmark of 400 BigCodeBench problems with controlled AST‑level corruptions, the authors quantify over‑editing and demonstrate that a simple preservation instruction can significantly reduce excess edits and cognitive complexity while improving Pass@1. They further explore post‑training strategies, finding that reinforcement learning yields the best balance between edit fidelity and performance retention, thereby establishing edit fidelity as a distinct, measurable dimension of code‑repair quality.

arXiv AI

Sep 4

When Models Edit Too Much: On the Fidelity of Minimal Code Edits

The paper investigates the problem of over‑editing by large language models when repairing code, showing that even state‑of‑the‑art models like GPT‑5.5 frequently rewrite more code than necessary. Using a benchmark of 400 BigCodeBench problems with controlled AST corruptions, the authors quantify excess edits and demonstrate that a simple preservation instruction can reduce unnecessary changes and improve pass rates. They further explore training strategies, finding that reinforcement learning yields the best balance between edit fidelity and performance retention, highlighting edit fidelity as a distinct, measurable dimension of code‑repair quality.

By Tongyao Zhu, Wei Hern Lim, Min-Yen Kan

llms reinforcement-learning fine-tuning

arXiv AI

Jun 30

Can Fine-Tuning Erase Your Edits? On the Fragile Coexistence of Knowledge Editing and Adaptation

arXiv:2511. 05852v4 Announce Type: replace-cross Abstract: Knowledge editing (KE) offers a lightweight alternative to retraining for updating large language models (LLMs).

By Yinjie Cheng, Paul Youssef, Christin Seifert, J\"org Schl\"otterer, Zhixue Zhao

llms fine-tuning benchmarks safety

arXiv AI

Aug 3

RePaCA: Leveraging Reasoning Large Language Models for Static Automated Patch Correctness Assessment

arXiv:2507. 22580v2 Announce Type: replace-cross Abstract: Automated Program Repair (APR) seeks to automatically correct software bugs without requiring human intervention.

By Marcos Fuster-Pena, David de-Fitero-Dominguez, Antonio Garcia-Cabot, Eva Garcia-Lopez

llms reinforcement-learning benchmarks

arXiv AI

Aug 3

To Add Is Machine, To Delete Is Human: Measuring and Mitigating Deletion Avoidance in LLM Code Editing

arXiv:2607. 28887v1 Announce Type: cross Abstract: Large language models increasingly write and repair production code, yet evidence is mounting that their test-passing patches leave codebases harder to maintain.

By Amir M. Ebrahimi, Mohammed Mehedi Hasan, Aaditya Bhatia, Gopi Krishnan Rajbahadur, Ahmed E. Hassan

llms benchmarks

arXiv Machine Learning

Jun 26

Reproducibility Study of "AlphaEdit: Null-Space Constrained Knowledge Editing for Language Models"

arXiv:2606. 26783v1 Announce Type: new Abstract: Fang et al.

By Ananth K S, Arya Hariharan

llms benchmarks safety

arXiv AI

1d ago

Suppressed, Not Erased: A Representational Trace of Edited Facts Survives Even Weight-Free Knowledge Editing

The paper investigates whether knowledge editing truly erases original facts from language models. Using a linear trace probe, the authors find that after editing a fact in GPT‑2‑XL, the original object remains highly decodable from hidden states across three different editing methods, even when the model behaves correctly on edited prompts. This suggests that editing suppresses rather than removes the original association in representational space.

By Priyansh Srivastava, Romit Chatterjee

llms fine-tuning benchmarks

arXiv AI

Sep 10

Diffs vs. Whole Files: An Empirical Comparison of Iterative Edit-Based and Direct Generation for Flutter/Dart Code Models

The paper compares two output regimes for code-editing language models: direct generation, where the model outputs the entire modified file, and iterative diff-based generation, where the model emits a sequence of localized edits. Experiments on Flutter/Dart tasks show that direct generation consistently outperforms diff-based generation across metrics such as compilation success, token efficiency, and quality judgments. However, diff-based generation can be competitive for short, spatially localized edits, particularly in refactoring and error-handling tasks with few edit steps.

By Andrej Andrejev

llms fine-tuning

arXiv AI

Jul 29

Learning from 53.6K Real-World Developer Edits of AI-Generated Code

arXiv:2607. 25130v1 Announce Type: cross Abstract: Imperfections in AI-generated code require that software developers modify the generated code manually, or by re-prompting an AI programming assistant.

By Jenny T. Liang, Mihika Bairathi, Wayne Chi, Ameet Talwalkar, Nishant Subramani, Valerie Chen

llms fine-tuning benchmarks

arXiv AI

Jul 7

Obey, Diverge, Collapse: Blind Obedience to Incorrect Instructions Drives Code LLMs to Irrecoverable Code Semantic Collapse

arXiv:2607. 04537v1 Announce Type: cross Abstract: Code language models are now trusted collaborators in production workflows for debugging, refactoring, and iterative repair, and every benchmark that evaluates them assumes the instructions they act on are correct.

By Raj Jaiswal, Anany Singh Divy, Savar Bhasin, Adi Bajpai, Tanuja Ganu, Rajiv Ratn Shah

llms benchmarks

arXiv AI

3d ago

Efficiency Hallucination: Formalizing and Measuring Behavioral Calibration in LLM-Based Code Optimization

arXiv:2609.14839v1 Announce Type: cross Abstract: The integration of Large Language Models (LLMs) into automated code optimization introduces a critical reliability risk we term the Efficiency Halluc...

By Sarah Wilson, Gail Kaiser, Patrick Musau

llms benchmarks safety

Hugging Face Trending Papers

Jul 5

Obey, Diverge, Collapse: Blind Obedience to Incorrect Instructions Drives Code LLMs to Irrecoverable Code Semantic Collapse

Code language models are now trusted collaborators in production workflows for debugging, refactoring, and iterative repair, and every benchmark that evaluates them assumes the instructions they act on are correct. We study what happens when that assumption breaks.

llms benchmarks

arXiv Computation and Language

Sep 4

CROCODIL: Cross-Model Code Editing with LLMs

CROCODIL is a post‑training framework designed to improve cross‑model code editing with large language models (LLMs). It addresses the problem that different LLMs, trained on distinct datasets, often make excessive edits when applied to code generated by another model. By combining a similarity reward that discourages large changes with an execution reward that ensures build and test success, CROCODIL encourages smaller, functionally correct edits.

By Linghan Zhong, Aditya Thimmaiah, Jayanth Srinivasa, Milos Gligoric, Junyi Jessy Li

llms

Related stories